It is one of the happy incidents of the federal system, Justice Louis Brandeis once mused, that a single courageous state may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country. Well, quite. These days we practically expect the states to try their hand at fixing tricky national problems before the federal government steps in. So to many observers, when several states recently turned their attention to providing health care for the uninsuredone of the thorniest domestic problems of allit looked like cause for considerable optimism.

For a start, both Massachusetts (which passed a near-universal plan last year) and California (which is seriously debating one) are trying something relatively new: individual mandates that require every person in the state to have health insurance. The fact that Democratic legislatures have pushed Republican governorsone a presidential candidate (Mitt Romney), the other a celebrity (Arnold Schwarzenegger)to back the notion of universal care has given this per-ennial liberal dream a bipartisan cachet. It has also helped revive the national conversation around universal health care by moving the discussion from airy, moral exhortations to practical examples of possible paths forward. In fact, the leading Democratic presidential candidates have all committed themselves to universal health reform, some with plans that seem more than a little influenced by those statewide R&D centers. The laboratories of democracy appear to be in fine working order.

So far, so good. But we know that the politics of health care is treacherous. In the last half century, two presidents (Truman and Clinton) mounted serious efforts to provide health care to all the uninsured and failed, while another (FDR) scotched his ambitions when he appraised the obstacles. Theres no reason to think that things will be any easier this time around. Consider this: congressional Democrats, despite their new majority, were recently unable to pass even minor tweaks to the Medicare prescription drug benefit. If a Democrat wins the presidency in 2008, he or she will face a daunting array of moneyed interests aiming to kill any universal health care program they attempt to pass.

Faced with such long odds, a conventional wisdom will soon develop that it would be better for Washington to defer efforts to pass universal health care, perhaps indefinitely, and instead just leave the problem to the states to figure out. Indeed, such arguments are already being voiced by ideological opponents of expanded federal power. Lets just try many different approaches in many different states and see what happens, the Heritage Foundations Stuart Butler told me. The more complex an issue is, the less possible it is to actually know the best system in advancetherefore a system of trial and error in which all these pieces work together is the best way to push forward.

The idea of giving universal health care a little more time in the laboratories of democracy may sound tempting to certain cautious, bipartisanship-loving Beltway observers. But letting states continue to take the lead would be disastrous, for one very simple reason: providing health care for all citizens is one of those tasks, like national defense, that the states are simply unequipped to manage on their own. The history of state health reform initiatives (and theres quite a history) is a tale of false hopes and great disappointments. The deck is stacked from the start, and the housein this case the insurers, the providers, and other agents of the status quoalways wins. The new raft of reforms may prove different, but they probably wont. Universal care advocates must be realistic about that, and think hard about how to convert the energy in the states into a national solution before the current crop of novel experiments failbecause fail they almost certainly will.

The current appetite for universal health care in state capitals may seem thrilling and unprecedented to some, but to those who follow the issue it carries an unsettling charge of dj vu. Over the years, states have tried programs of many different ideological and economic persuasions. All of them failed, and not because the programs were insufficiently inventive, but because states are structurally incapable of sustaining them.

Washington State, for instance, opted for a Clintonian-style program. In 1993, it passed the Washington Health Services Act, which sought to tame the excesses of health markets and use the states regulatory powers to mandate universal coverage, force insurers to offer comprehensive benefit packages at affordable prices, and pool purchasing power so all residents could buy in. Because the insurers would rebel against this scheme to rip apart their business model and end their ability to price out the sick while attracting the well, the state demanded that all insurers who wanted access to Oregons market become certified through the statethat way, they couldnt undercut the program. But the insurers took a harder line than expected and instantly rebelled against the arrangement; many went so far as to cease doing business in the state. Others instantly jacked up premiums, creating total chaos. The next year, the Republican revolution washed over the statehouse, and by 1995 the new GOP majority had repealed most of the plans elements, including the parts forcing insurers to participate. By 2000, not a single insurer remained in the now-voluntary system. Today, 13 percent of Washingtons population is uninsured, compared with just over 10 percent in 1992. The state spends 12 percent of its gross state product on health care, just 1 percent less than the national average.

Hawaii, in contrast, went for the classic liberal approach: a giant government program. The state had already forced all employers to cover their workers in 1974. In 1994, hoping to cover all those without employer-provided insurance, Hawaii created the QUEST program, into which it merged Medicaid and a similar state program. Early reviews were good. At the outset, QUEST appeared so successful that Hillary Clinton used it as supporting evidence for the Clinton Health Security Act, proclaiming that Hawaii has achieved nearly universal coverage and has less of a cost.

What happened next wasnt pretty. First, Hawaii suffered from the problem known as the woodwork effect, where a new government program entices people out of the woodwork, leading to more applicants and higher costs than anticipated. An economic downturn proved even more damaging. Its a cruel economic ironybut an inescapable policy realitythat recessions rob government of the revenue it needs to cover the uninsured at precisely the moment that the most people need subsidies to get them through the lean times. And states are incapable of responding, since they, unlike the federal government, are constitutionally barred from running deficits.

So Hawaii responded in the only way it could: by cutting back the program. Instead of offering benefits for those whose income was up to three times the federal poverty line, it restricted eligibility to people whose income was twice the poverty line, and introduced an assets test. The dreams of universality evaporated. According to the Kaiser Family Foundation, 9 percent of Hawaiis population is uninsured, compared to 15 percent nationwide, and it spends 12 percent of its gross state product on health careexactly the same proportion as the rest of the country.

Tennessees health reform sprang from more conservative principles. In 1994, in the thick of the Republican revolution, Tennessee replaced its Medicaid program with a new plan for low-income residents called TennCare. TennCare was part of the managed care movement of the 1990s, an attempt to bring some free market discipline to the apparent bloat of government entitlements. It replaced the traditional fee-for-service structures of Medicaid with a new system wherein health maintenance organizations were given a fixed fee per patient, per month, thus theoretically creating incentives to keep the patients well rather than prescribe ever more treatments. It was ambitiousseeking to cover the poor, the young, the uninsured, and even the uninsurable (mainly those with expensive or chronic medical conditions). The New York Times called it for better and worse, the most sweeping revolution of a government-financed health care system in the nation.

Like Hawaiis experiment, TennCare seemed to work, at first. A few years after its inception, it covered a quarter of Tennessees population. But it ultimately met with exactly the same fate as Hawaiis program. A fiscal downturn created greater need and insufficient tax revenues. Costs increased. When voters rejected a tax hike, the state began cutting benefits. By 2005, Democratic Governor Phil Bredesen had dissolved much of the program, leaving more than 300,000 Tennesseans without coverage. I say to you with a clear heart, said Bredesen, that Ive tried everything. There is no big lump of federal money that will make the problem go away.

Finally, in the early 1990s, Oregon adopted a plan that defied ideological pigeonholing but was notable for attempting to do the one thing that really needs doing: controlling costs. Reformers fastened on an expansion of Medicaid as the quickest route to expanded access, and vowed to finance the expansion not with new funds, but with savings obtained through a rationing system that ranked treatments by cost-effectiveness and only covered those with the greatest proven benefits. By increasing access without depending on unstable streams of new revenues, Oregons plan sought to protect itself from the fiscal downturns and strains that bedeviled other efforts. The New York Times dubbed it a brave medical experiment. Between 1992 and 1996, the states uninsured population dropped by 7 percent, and the electorate voted twice to hike cigarette taxes in order to better fund the program.

In 2002, Governor John Kitzhaber decided to push Oregon closer to full coverage. However, the state was undergoing an economic downturn, and had no extra revenues to spare. So Kitzhaber decided to cover more people by cutting services and increasing the amount beneficiaries would have to spend out-of-pocket. This trade-off allowed the plan to achieve bipartisan support, but also caused it to tank spectacularly. Instead of expanding by 50 percent, the program contracted by 75 percent, as the increased cost sharing and stringent payment regulations caused thousands of Oregonians to leave or be kicked off the states rolls.

The new crop of state health care plans are trying to avoid the failures of the past by attempting something new: individual mandates. Here, Massachusetts led the way. Its plan forces everyone in the state to have health insurance, with subsidies for those who cant afford it. The genius of this approach is that many of those who lack health insurance arent poor or ailingtheyre middle-class people who have decided theyre healthy enough that they dont need to spend money on health coverage. Forcing these people to buy insurance (like similar reforms that require all drivers to have auto insurance) has the useful effect of bringing more money into the system while spreading the risk. So far, Massachusetts is the first state to pass such a plan, but other states, including California, Illinois, New Jersey, and Connecticut, are contemplating universal plans with similar individual mandates.

And the early indicators from Massachusetts suggest that the new program might actually workat least in Massachusetts. State officials recently announced that they expect to be able to afford sliding-scale premiums that would provide coveragesome of it bare bones, but still better than nothingto 99 percent of adults in the state.

Alas, Massachusetts may be the exception that proves the rule, for several reasons. First, it has one of the lowest uninsured populations in the country: 10 percent compared to the nationwide average of 16 percent. Second, a relatively large portion of that 10 percent can afford to buy insurance without government subsidy, thanks to the fact that the state is one of the nations wealthiest, with a median household income of about $8,000 above the national average. That leaves Massachusetts with a reasonably small number of citizens who need government support to purchase health insurance. And in that regard, Massachusetts has a third advantage: a ready-made funding source. In 1988, then Governor Michael Dukakis signed a pay-or-play universal health care bill into law. Like similar laws enacted in other states, it failed, and most of its provisions were repealed a year later by his successor. But portions of it remained, including revenue streamstotaling $540 milliondedicated to reimbursing hospitals who care for the uninsured. Thats $540 million that can now go to covering the uninsured. Its also money most other states dont have just lying around.

For those reasons and others, its unlikely that any other state can succeed at what Massachusetts is doing. Thats especially true of poor states, like New Mexico, which has 26 percent uninsured and a median household income thats 20 percent lower than the national average. Even if everyone in the state without insurance were mandated to have it, only a relatively small portion could afford to buy that insurance without massively generous subsidies. To provide those subsidies, the government would have to seek tax increases of a size that no state in recent memory has been able to pass. And to sustain the program, it would have to hold firm during all manner of recessions and downturns, when the calls for tax relief grow strongest but when the programs funding may actually need to be increased.

Even states with higher household incomes have only a limited ability to fund a universal health care scheme. In Illinois, for instance, the individual mandate plan that Governor Rod Blagojevich has proposed wont be enough to cover the programs cost. So he has suggested a tax on gross receipts on all business of revenue of $1 million a year or more. The policy is widely seen as putting Illinois businesses at an onerous disadvantage compared to other statesand not just by the usual suspects. The Reverend Jesse Jackson, not exactly a die-hard supply-sider, has called it a bad idea. When youve managed to put Jesse Jackson and the Chamber of Commerce on the same side, youve lost pretty much everyone.

Similarly, in California individual mandates wont cover the cost of Schwarzeneggers much-touted universal health care plan. He plans to lobby the federal government for $3.7 billion in annual subsidies. Its an odd expectation, given that President Bushs 2006 budget called for $4.6 billion in health care cuts to the state over the next decade. But lets assume that Schwarzenegger gets his money. What if, say, John McCain is then elected president in 2008 and is able to follow through on his constant promises to shrink the size of government, and cuts California loose? Or what if the economy takes a dive, and Congress scales back contributions? Thats the end of Californias program.

One of the great things about state governments is that they have more freedom than the federal government does to test new policy ideas. But it pays to look honestly at what the results of those tests actually say. And in this case, the results are pretty clear: states are no good at delivering universal health care.

No one can doubt the role Massachusetts and California have played in reinvigorating the debate over national health care. And if the reforms currently percolating at the state level help provide momentum for a national health care system in the next few years, all the effort will have been worth it. If they dont, however, they may ultimately prove detrimental. If high-profile efforts like those in Massachusetts and California cant be properly implemented, or are launched and then collapse, theyll become powerful weapons in the hands of protectors of the status quo. After the demise of Washington States plan, for instance, the Heritage Foundation published an article stating that the program gave state legislators around the country an experimental taste of how a Clinton-style health care plan would workor fail to work. The result was higher costs, burgeoning bureaucracy, and micromanagement. If the example of Washington is replaced with California, this kind of attack will become far more deadly.

The rest of the industrialized world proves that nationally run care can, and does, work. But our own history proves that state-run care doesnt, and reformers should keep that in mind. Brandeis, for all his federalism, was a realist, too. If we would guide by the light of reason, he said, we must let our minds be bold. On health reform, the light of reason is clear. We must merely be bold enough to follow it, and not settle for smaller, unsustainable victories because we fear the battle necessary for an enduring triumph.

Our ideas can save democracy... But we need your help! Donate Now!

Ezra Klein

Follow Ezra on Twitter @ezraklein. Ezra Klein is the founder and editor-in-chief of